AI, AI, oh!
There’s a new exhibit opening at the San Antonio Art League & Museum next week called Lone Star Horizons: Texas Landscapes Past, Present Future. I have two pieces in it – one that was juried in, and one that I created for a special section about artists who use Artificial Intelligence tools in various ways in their work. There are nine other artists in this section – stay tuned and I’ll show them to you once the show opens.
Images generated with AI tools fascinate me, not because it want to find an easy way to make a picture, but because the tool itself has so many possibilities — and limits. For example, I created a series of coordinated collage mages in my Sisters workshop that are designed to teach combinations and composition without making students worry about copyright images from commercial sources. To me, that’s a help
It’s also fun to test the way that AI applications like DALL-E monitor themselves regarding content. As an experiment, I asked DALL-E, which is an image generator that make pictures from scratch, to create this:
Oops – it seems that “in the style of Georgia O’Keeffe” is not allowed. This makes me feel good because it shows level of protection for that artist. You’ll notice that DALL-E changed its “mind” before could generate the image.
Look what happens when I change the prompt and leave out the painter’s name:
Now it produces an image, one that is completely new and not copyright. Interesting!
But when I tried another prompt and asked simply for the “style of southwestern painters” it would not generate it.
When I removed the “style” part, look what happened:
It’s good to know that there are some limits. So how does this help with the creative process?
AI is a visualization tool for me. Let’s say I am starting on a new series of Spirit Dolls. I can use AI to generate a concept or design that gives me ideas on directions to follow even before I get out the fiber and clay. Here’s an example:
Obviously, none of these generated images goes with my signature styles, but I can try out the idea with Photoshop, using one of my Earthshard faces.
It’s a nice graphic, and while I like some of it, I can eliminate most of it as a possibility (like too much silver) without having to actually build it than take it apart again.
By the way, here is the AI-aided piece that will be in the Texas Landscape show. It is an encaustic collage with parts generated by the AI tool called MidJourney, and some parts from other sources, including some real pressed and dried flowers. It’s covered with layers of encaustic medium. And no Georgia O’Keeffe paintings were swiped in the production of this piece! 🙂
To me, a piece like this represents a fusion of old and new tools – an experiment that plays with new frontiers that we can chose to use or not. We follow out hearts and can go back and forth between comfort and experimentation.
No matter what the situation, remind yourself, “I Have A Choice.” -Deepak Chopra
Thanks for reading!
PS If you wold like copies of any of the images that were generated in the examples, just let me know and I will make them available for download. They are not copyrighted and you can use them in any way you wish as long as you don’t say that you painted them! LOL.
Fascinating story……Thanks so much for sharing the details. Darlene
It really is interesting stuff – 🙂
Thank you Lyn….
What a great idea to try. I’ve been curious about AI for quite some time. This seems like a great idea for starting something grand. Thank you much.
I’d like to try this with my grandsons. They are very creative and I want to feed that.